AWS DeepLens
Developer Guide

Getting Started with AWS DeepLens

Before building a computer vision application project and then deploying the project to your AWS DeepLens device, you must first accomplish the following:

  1. Set up your AWS DeepLens application development environment:

    For this task, you sign up for an AWS account, if you have not already got one, You should also create an IAM user for building and deploying your AWS DeepLens project. Additionally, you should have some understanding of the permissions required to use AWS DeepLens. You can leverage the AWS DeepLens console to define the required permissions for you. Alternatively, you can use the IAM console or call the IAM API to define them yourself.

  2. Register your AWS DeepLens device:

    For this task, you name your device for the AWS DeepLens service to identify it; grant IAM permissions for you to create and deploy an AWS DeepLens project; call AWS IoT to generate a security certificate for the device; and request AWS IoT Greengrass to create an AWS IoT thing representation for your device. You do these on the AWS Cloud. To complete the registration, you must call the setup app on the device after connecting your computer to the device's local Wi-Fi network, also known as its AMDC-NNNN) network. To set up the device, you choose to use a home or office network to connect to the internet, upload the security certificate generated by AWS IoT for the device, and set the password for device logins.

  3. Verify the device registration status.

    You can deploy your AWS DeepLens project to your device only when the device is successfully registered and online. Here, online means that the device is connected to the Internet and authenticated by the AWS Cloud. Hence, verifying the registration status amounts to ensuring that the internet connection remains open and the correct security certificate is uploaded to the device.

After you've successfully registered your AWS DeepLens device and connected the device to the AWS cloud, you typically proceed to explore AWS DeepLens by performing the following:

  1. Building an AWS DeepLens computer vision application project, including training a deep learning computer vision model and developing an inference Lambda function.

  2. Deploying the project to run on the device.

  3. Inspecting the project output to verify the inference result.

To introduce you to this process end-to-end, we will walk you through all the above-mentioned tasks in this section, focusing on AWS DeepLens development environment setup and the AWS DeepLens device registration, while highlighting project creation, deployment and result viewing with a sample project containing a pre-trained model and a tested inference Lambda function.